“CAE vendors spent a lot of time trying to bring CAD into the simulation process,” said Joe Walsh of intrinSIM. “Now they are trying to find a way to take CAD out.”
Walsh is addressing the over 300 assembled at NAFEMS 2016 in Lynnwood, Wash., north of Seattle. The once lumber mill town is now Boeing territory. Boeing employs 35,000 people in its nearby Everett facility.
The idea of cutting CAD and CAM out of the process is an idea Walsh credits to Andreas Vlahinos, who spoke about optimized lattice structures at COFES. If a part’s shape can be optimized using optimization software (essentially a CAE program) and then 3D printed, there’s no need for it to be modeled in CAD and manufactured with CAM. The part goes right from CAE to be 3D printed. It’s a short circuit.
We’re not quite there yet, warned Walsh. A lot of things have to be in place before shapes can be optimized automatically and 3D printed at the push of a button. For one, topology optimization has to be easier to use, widely available and robust. It would help if 3D printing was indeed push-button easy—it isn’t. But a small, yet growing, number of parts are currently complex enough to fit the shape-to-part bill, said Walsh. In fact, for some parts, the only way they can possibly be manufactured is if CAD and CAM don’t get in the way.
Take for example a part that, for purposes of saving material, is composed of a lightweight lattice structure. You might have a chance of designing it in CAD if the lattice is repeatable and of uniform size. Good luck if you want it to change orientation and size to accommodate variations in shape and loading. And even if that can be painstakingly designed, is there any chance at all of being able to program it on the CNC machine?
CAD and CAM are powerless to design or manufacture such a complex geometry. However, topology optimization produces such organic shapes by its very nature. And 3D printing cares little of the complexity of the shape.
A variable lattice is just as easy for it to make as an ordinary part.